Saffron Agronomy and Technology

Abstract Saffron is a triploid, sterile plant, used as a spice and medicinalplant in all countries. Stigma is the most important part of saffron. So far no reliable molecular methods were provided to identify and prediction of the three/multi branches species. In this study, using different bioinformatics algorithms, new tools for prediction based on Sequence-Related Amplified Polymorphismmolecular markers is presented. Five alleles M1311400, M151200, M12100 and M10850 selected as the most important classifier by Attribute Weighting models which has the potential to cluster and recognize the three from multi branches stigma. K-Means and K-Medoids unsupervised clustering algorithms were fully able to cluster each genotype to the right classes. Our results showed that for the first time, data mining techniques can be effectively used to genetic differentiation between three and multi stigma with above 90 percent the accuracy andprecision. These methods can use in gene mapping and selection by biomarker.